Robot motion data generation system, motion data generation method, and storage medium

Abstract

This application relates to a motion data generation system, a motion data generation method, and a storage medium for a robot. A motion data generation system for the robot is provided. The robot includes an upper body, a waist, and a lower body. The motion data generation system includes a subject motion data acquisition unit that acquires upper body motion data captured from the motion of the upper body of the subject and waist motion data captured from the motion of the waist of the subject; a manually generated motion data acquisition unit that acquires lower body motion data and manually generated waist motion data, the lower body motion data and the manually generated waist motion data being generated by manual input of a user; and a robot motion control unit. The robot motion control unit includes a leg state determination unit and a waist motion data generation unit.

Description

Technical Field

[0001] The present application relates to a motion data generation system, a motion data generation method and a storage medium for a robot, and in particular to a motion data generation system, a motion data generation method and a storage medium for a humanoid robot. Background Art

[0002] In a motion data generation system disclosed in Japanese Unexamined Patent Application Publication No. 2012-223864 (JP 2012-223864 A), motion data is acquired from the motion of a subject, and a humanoid robot can operate based on the motion data. Summary of the invention

[0003] The inventors of the present application have found the following problems.

[0004] Depending on the part of the robot, the robot may be required to operate at its maximum performance. In the motion data generated by the motion data generation system, the robot's motion is limited by the subject's motion. Therefore, depending on the part of the robot, it may not be possible to achieve a motion in which the robot's own performance is fully demonstrated.

[0005] The present invention is made in view of the above-mentioned problems, and an object of the present invention is to provide a robot motion data generation system, a motion data generation method and a storage medium, which generate motion data capable of performing actions in which the performance of the robot itself is fully demonstrated.

[0006] The motion data generation system of a robot according to the present invention is a motion data generation system for a robot including an upper body, a waist, and a lower body. The motion data generation system includes a subject motion data acquisition unit that acquires upper body motion data captured from the motion of the upper body of the subject and waist motion data captured from the motion of the waist of the subject; a manually generated motion data acquisition unit that acquires lower body motion data and manually generated waist motion data, the lower body motion data and the manually generated waist motion data being generated by manual input of a user; and a robot motion control unit that includes a leg state determination unit and a waist motion data generation unit. The leg state determination unit determines whether at least one leg of the robot is in a swinging state or both legs of the robot are in a standing state. When the leg state determination unit determines that at least one leg of the robot is in the swinging state, the waist motion data generation unit generates the manually generated waist motion data as the robot waist motion data. When the leg state determination unit determines that both legs of the robot are in the standing state, the waist motion data generation unit generates the captured waist motion data as the robot waist motion data. The robot motion control unit controls the upper body, the waist, and the lower body of the robot based on the upper body motion data, the robot waist motion data, and the lower body motion data, respectively.

[0007] According to this configuration, when at least one leg of the robot is in a swinging state, the waist of the robot is operated based on the manually generated waist motion data. Therefore, the robot is not likely to lose balance and the robot can be operated according to its own performance.

[0008] During a transition period from a standing phase in which both legs of the robot are in the standing state to a swinging phase in which at least one leg of the robot is in the swinging state, the waist motion data generation unit may generate the robot waist motion data by synthesizing the manually generated waist motion data and the captured waist motion data.

[0009] According to this configuration, the robot waist motion data during the transition period is synthesized data of the manually generated waist motion data and the captured waist motion data. Therefore, a significant displacement of the waist position of the robot relative to the upper body and the lower body of the robot is suppressed. That is, an unnatural motion of the waist of the robot is suppressed.

[0010] From the start to the end of the transition period, in the robot waist motion data generated from the synthesized data, a reference ratio referring to the manually generated waist motion data may gradually increase with respect to the captured waist motion data.

[0011] According to this configuration, during the transition period, it is possible to transition to the swing phase while ensuring the characteristics of the waist movement based on the captured waist movement data. Therefore, significant displacement of the waist position of the robot relative to the upper body and the lower body of the robot is suppressed.

[0012] The method for generating motion data according to the present invention is a method for generating motion data of a robot executed in a motion data generation system of a robot including an upper body, a waist, and a lower body. The method for generating motion data includes steps of acquiring upper body motion data captured from the motion of the upper body of a subject and captured waist motion data captured from the motion of the waist of the subject, acquiring lower body motion data and manually generated waist motion data, the lower body motion data and the manually generated waist motion data being generated by manual input of a user, a step of determining whether at least one leg of the robot is in a swing state or whether both legs of the robot are in a standing state, when it is determined that at least one leg of the robot is in the swing state, generating the manually generated waist motion data as the waist motion data of the robot, and when it is determined that both legs of the robot are in the standing state, generating the captured waist motion data as the waist motion data of the robot, and a step of controlling the upper body, the waist, and the lower body of the robot based on the upper body motion data, the waist motion data of the robot, and the lower body motion data, respectively.

[0013] According to this configuration, when at least one leg of the robot is in a swing state, the waist of the robot is operated based on the manually generated waist motion data. Therefore, the robot is not easily out of balance and can operate the robot according to its own performance.

[0014] The storage medium according to the present invention stores a robot motion data generation program executed by a computer that operates an arithmetic device in a robot motion data generation system. The robot includes an upper body, a waist, and a lower body. The motion data generation program causes the computer to execute: steps of acquiring upper body motion data captured from the motion of the upper body of a subject and waist motion data captured from the motion of the waist of the subject, steps of acquiring lower body motion data and manually generated waist motion data, where the lower body motion data and the manually generated waist motion data are generated by manual input of a user, a step of determining whether at least one leg of the robot is in a swinging state or whether both legs of the robot are in a standing state, when it is determined that at least one leg of the robot is in the swinging state, generating the manually generated waist motion data as robot waist motion data, and when it is determined that both legs of the robot are in the standing state, generating the captured waist motion data as the robot waist motion data, and steps of controlling the upper body, the waist, and the lower body of the robot based on the upper body motion data, the robot waist motion data, and the lower body motion data respectively.

[0015] According to this configuration, when at least one leg of the robot is in a swinging state, the waist of the robot is operated based on the manually generated waist motion data. Therefore, the robot is not easily unbalanced and can operate the robot according to the performance of the robot itself.

[0016] The present application can provide a robot motion data generation system, a motion data generation method, and a storage medium that generate motion data capable of performing motions that fully exhibit the performance of the robot itself. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Hereinafter, features, advantages, and technical and industrial significance of exemplary embodiments of the present invention will be described with reference to the drawings, in which the same reference numerals denote the same elements, and in which:

[0018] Figure 1 is a schematic diagram showing a configuration example of a robot controlled by a motion data generation system according to a first embodiment;

[0019] Figure 2 is a block diagram showing a schematic system configuration of a motion data generation system according to a first embodiment;

[0020] Figure 3 is a block diagram showing a schematic system configuration of an arithmetic device of a motion data generation system according to a first embodiment;

[0021] Figure 4is a diagram showing the positions of markers attached to a subject and the states in which the positions of the respective markers have been repositioned for a small humanoid robot or a long-legged humanoid robot;

[0022] Figure 5 is a diagram showing an example of robot waist motion data in the motion data generation system according to the first embodiment; and

[0023] Figure 6 is a flowchart showing an example of the operation of the motion data generation system according to the first embodiment. DETAILED DESCRIPTION

[0024] Hereinafter, specific embodiments of the present invention will be described in detail with reference to the accompanying drawings. However, the present invention is not limited to the following embodiments. In addition, for the sake of clarity, the following description and drawings have been appropriately simplified.

[0025] First Embodiment

[0026] will be described with reference to Figures 1 to 4 the first embodiment. Figure 1 is a schematic diagram showing a configuration example of a robot controlled by a motion data generation system according to the first embodiment.

[0027] As Figure 1 shown, the robot 100 includes a lower body 103, a waist 102, and an upper body 101.

[0028] The lower body 103 may include at least two legs. The lower body 103 according to the first embodiment has a structure corresponding to each of a human thigh, calf, and foot. The lower body 103 supports the waist 102 and the upper body 101. The lower body 103 may include two legs, but may include three or more legs.

[0029] The waist 102 connects the lower body 103 and the upper body 101. For example, the waist 102 changes the posture. The posture includes the three-dimensional position and rotation of the joints.

[0030] The upper body 101 may have a structure corresponding to at least one of a human head, neck, torso, arms, hands, and fingers. The upper body 101 according to the first embodiment has a structure corresponding to each of a human head, torso, arms, and hands.

[0031] Figure 2It is a block diagram showing a schematic system configuration of an action data generation system according to a first embodiment. The action data generation system 10 according to the first embodiment includes an action capturing device 1 and an arithmetic device 2. The action capturing device 1 acquires action data from the actions of a subject. The arithmetic device 2 generates action data of a humanoid robot (anthropomorphic robot) such as a bipedal walking robot based on the action data acquired by the action capturing device 1. The action data generation system 10 can generate natural action data of a humanoid robot that is closer to the actions of the subject.

[0032] The action capturing device 1 can be any device that acquires the positions of the respective joints of the upper body and waist of the subject. The action capturing device 1 can also acquire the angles of the respective joints of the upper body and waist of the subject. A tracker, a sensor, a camera image, etc. can be used for the action capturing device 1. The action capturing device 1 according to the first embodiment includes a plurality of markers 11, a tracker 12 for detecting the positions of the respective markers 11, a plurality of foot contact sensors 13, and a processing unit 14 for processing the actions of the respective markers 11 detected by the tracker 12 and the output signals (floor reaction force information, etc.) from the foot contact sensors 13.

[0033] Each marker 11 is attached to a focused part for measuring (capturing) the actions of the subject H1 (see Figure 4 ). The focused part includes at least the upper body and waist of the subject H1, and for example, includes the elbows, shoulders, head, arms, neck, hands, and fingers of the subject H1. The tracker 12 detects the positions of the respective markers 11 at a predetermined cycle, and the positions of the respective markers 11 detected by the tracker 12 are input to the processing unit 14. As described above, action data of the focused part of the subject is acquired. The processing unit 14 performs predetermined processing on the detected position data of the markers 11 and outputs the processed action data (action capture data) to the arithmetic device 12.

[0034] The arithmetic unit 2 generates natural humanoid robot action data that is closer to the actions of the subject based on the action data of the subject acquired by the action capturing device 1. Figure 3 It is a block diagram showing a schematic system configuration of the arithmetic device 2 of the action data generation system 10 according to the first embodiment. The arithmetic device 2 according to the first embodiment includes a scaling unit 21, a subject action data acquisition unit 22, a manually generated action data acquisition unit 23, and a robot action control unit 24.

[0035] The arithmetic unit 2 may also appropriately include at least one of the following: a data correction unit, a first inverse kinematics arithmetic unit, a target zero moment point (ZMP) calculation unit, a target center-of-gravity trajectory calculation unit, and a second inverse kinematics arithmetic unit. The data correction unit performs correction processing such as the ground contact state of the toes on the motion data on which the scaling unit 21 has performed retargeting processing. The first inverse kinematics arithmetic unit performs inverse kinematics arithmetic operations on the entire humanoid robot based on the motion data on which retargeting processing and correction processing have been performed, and calculates a sequence of joint angles (time-series data of each joint angle) of the humanoid robot. In addition, the first inverse kinematics arithmetic unit calculates the ZMP trajectory, center-of-gravity trajectory, angular momentum trajectory of the center of gravity, etc. of the humanoid robot before dynamic stabilization based on the calculated sequence of joint angles. The target ZMP calculation unit is a specific example of a target ZMP calculation device, and calculates a target ZMP trajectory for stabilizing the motion of the humanoid robot based on the ZMP trajectory calculated by the first inverse kinematics arithmetic unit. The target center-of-gravity trajectory calculation unit is a specific example of a target center-of-gravity trajectory calculation device, and calculates a target center-of-gravity trajectory based on the ZMP trajectory calculated by the first inverse kinematics arithmetic unit and the target ZMP trajectory calculated by the target ZMP calculation unit. The second inverse kinematics arithmetic unit is a specific example of a second inverse kinematics arithmetic device, and performs inverse kinematics arithmetic operations on the entire body of the humanoid robot based on the target center-of-gravity trajectory calculated by the target center-of-gravity trajectory calculation unit, and calculates a sequence of joint angles of the humanoid robot. The sequence of joint angles of the humanoid robot calculated in this way can be used as part of the motion data.

[0036] The arithmetic unit 2 is provided with a hardware configuration based on a microcomputer. The microcomputer includes, for example, a central processing unit (CPU) 2a for performing arithmetic processing and the like, a read-only memory (ROM) 2b in which an arithmetic program to be executed by the CPU 2a is stored, and a random access memory (RAM) 2c for temporarily storing processing data, etc. In addition, the CPU 2a, ROM 2b, and RAM 2c are connected to each other via a data bus 2d.

[0037] The scaling unit 21 is a specific example of a scaling device, and performs known retargeting processing on the motion data from the motion capture device 1 so that the motion data of the subject adapts to the actual operating humanoid robot. The motion data acquired by the motion capture device 1 is motion based on the lengths of the respective parts of the subject, and since information about the focal parts (for example, the positions and postures of the upper body and waist to be used as motion data and the angles of any joints) cannot be adapted to the humanoid robot as it is, retargeting processing is performed on the motion data.

[0038] For example, the scaling unit 21 determines the magnification ratio of each connection of the humanoid robot based on the ratio of the respective connection lengths of the humanoid robot to be applied to the lengths of the corresponding parts of the subject, and performs a repositioning process.

[0039] For example, as Figure 4 shown, in the case of performing a repositioning process on the positions of the respective markers 11 (the positions of the focal parts) attached to the subject H1, when the robot 100 as an example of the robot 100 is small, the respective markers 11 corresponding to the respective markers 11 in the motion data are close to each other. In addition, the robot 100b as an example of the robot 100 has long legs. Compared with the motion data of the robot 100a, the respective markers 11 corresponding to the respective markers 11 in the motion data of the robot 100b are separated from each other.

[0040] The subject motion data acquisition unit 22 acquires the motion data on which the scaling unit 21 has performed a repositioning process. The acquired motion data includes upper body motion data captured from the motion of the upper body of the subject H1 and waist motion data captured from the motion of the waist of the subject H1.

[0041] The manually generated motion data acquisition unit 23 acquires lower body motion data and manually generated waist motion data generated by a user's manual input. The lower body motion data and the manually generated waist motion data can be acquired from the ROM 2b of the arithmetic device 2, or can be acquired by a user's manual input via an interface or the like. The lower body motion data includes a gait pattern, a foot posture, and a posture of the waist 102 that establishes these.

[0042] The robot motion control unit 24 controls the upper body 101, the waist 102, and the lower body 103 of the robot 100. The robot motion control unit 24 includes a leg state determination unit 25 and a waist motion data generation unit 26.

[0043] Here, there is a swing phase in which at least one leg of the robot 100 is in a swing state while the robot 100 is working. In addition, there is a standing phase in which both legs of the robot 100 are in a standing state. In many cases, the swing phase and the standing phase alternate and repeat while the robot 100 is working.

[0044] The leg state determination unit 25 determines whether at least one leg of the robot 100 is in a swinging state or whether both legs of the robot 100 are in a standing state. Specifically, the leg state determination unit 25 can determine whether at least one leg of the robot 100 is in a swinging state based on the lower body motion data acquired by the manually generated motion data acquisition unit 23. More specifically, the leg state determination unit 25 determines whether at least one leg of the robot 100 is in a swinging state for each moment according to the moment in the lower body motion data. The leg state determination unit 25 determines whether at least one leg of the robot 100 is in a swinging state, and estimates whether at least one leg of the robot 100 is in a swinging phase or both legs of the robot 100 are in a standing phase at the current moment.

[0045] When the leg state determination unit 25 determines that at least one leg of the robot 100 is in a swinging state, the waist motion data generation unit 26 selects the manually generated waist motion data, and generates the manually generated waist motion data as the robot waist motion data. On the other hand, when the leg state determination unit 25 determines that both legs of the robot 100 are in a standing state, the waist motion data generation unit 26 selects the photographed waist motion data, and generates the photographed waist motion data as the robot waist motion data.

[0046] The robot motion control unit 24 operates the waist 102 of the robot 100 based on the robot waist motion data. In other words, the robot motion control unit 24 operates the waist 102 based on the manually generated waist motion data in the swinging phase. In addition, the robot motion control unit 24 operates the waist 102 based on the photographed waist motion data during the standing phase. During the transition period for transitioning from the standing phase to the swinging phase, the robot motion control unit 24 can smoothly transition the motion of the waist 102 from the motion based on the photographed waist motion data to the motion based on the manually generated waist motion data. The transition period can be set on the standing phase side.

[0047] Figure 5 is a diagram showing an example of the robot waist motion data in the motion data generation system according to the first embodiment. As Figure 5 shown, in principle, during the swinging phase, the waist 102 is operated based on the manually generated waist motion data. On the other hand, in principle, during the standing phase, the waist 102 is operated based on the photographed waist motion data. In addition, during the transition period, the waist 102 is operated so as to smoothly transition from the motion based on the photographed waist motion data to the motion based on the manually generated waist motion data.

[0048] The waist motion data generation unit 26 can generate robot waist motion data by synthesizing manually generated waist motion data and captured waist motion data. During the transition period from the standing phase to the swinging phase, the robot motion control unit 24 can operate the waist 102 based on the robot waist motion data generated from this synthesized data. As a result, it is possible to suppress a sudden change in the position of the waist 102 due to the displacement of the waist 102 of the robot 100 during the transition period. That is, unnatural motions of the waist 102 are suppressed.

[0049] As a specific example of the method for generating synthesized data, there is a method for generating synthesized data in which, during the transition period, the reference ratio (hereinafter referred to as the reference ratio) for referring to the captured waist motion data or the manually generated waist motion data increases or decreases with respect to the position of the waist 102. The start time of the transition period is set as the transition start time t s , and the end time of the transition period is set as the transition end time t e . Specifically, at the transition start time t s， , the reference ratio for referring to the captured waist motion data is set to be higher than that of the manually generated waist motion data. From the transition start time t s to the transition end time t e , the reference ratio for referring to the manually generated waist motion data gradually increases with respect to the captured waist motion data. In a specific example of the method for generating synthesized data, during the transition period of the synthesized data, it is possible to transition to the swinging phase while maintaining to some extent the motion characteristics of the waist 102 based on the captured waist motion data.

[0050] In a specific example of the method for generating synthesized data, linear interpolation can be used. As a specific example of linear interpolation, there is a method for obtaining the posture Pt of the waist 102 of the robot 100 by using the following formula (1). In other words, the following formula (1) shows the relationship between the posture Pt of the waist 102 of the robot 100, the transition start time t s , the transition end time t e , the posture Pht of the waist 102 based on the manually generated waist motion data, and the posture Pct of the waist 102 based on the captured waist motion data.

[0051] [Formula 1]

[0052]

[0053] When switching from the swing phase to the stance phase, the robot motion control unit 24 obtains the posture Prt of the waist 102 using the motion of the waist 102 based on manually generated waist motion data as a reference. In this case, the posture Prt of the waist 102 at time t, the switching time t for switching from the swing phase to the stance phase are shown by the following formula (2). c The posture Prt of the waist 102 at time t c , a constant A for normalizing the waist posture of the subject H1 to the size of the robot 100, the posture Pht of the waist 102 based on the captured waist motion data at time t, and the posture Pht of the waist 102 based on the captured motion data at the switching time t c The posture Pht of the waist 102 at time t c The relationship between them. That is, the posture Prt of the waist 102 can be obtained by using the following formula (2).

[0054] [Formula 2]

[0055] Prt = Prt c + A(Pht - Pht c )…(2)

[0056] The robot motion control unit 24 generates motion data by combining the upper body motion data acquired by the subject motion data acquisition unit 22, the lower body motion data acquired by the manually generated motion data acquisition unit 23, the manually generated waist motion data, the captured waist motion data, and the synthesized data.

[0057] In addition, the robot motion control unit 24 can perform stabilization processing on the front-back and left-right direction components of the robot 100 in the posture Prt of the waist 102 according to the target ZMP trajectory to correct the motion data. As described above, when the arithmetic device 2 includes a data correction unit, a first inverse kinematics arithmetic unit, a target ZMP calculation unit, etc., the target ZMP trajectory can be obtained. The robot motion control unit 24 generates a control signal based on the motion data and sends the control signal to the upper body 101, the waist 102, and the lower body 103 of the robot 100. Based on this control signal, the upper body 101, the waist 102, and the lower body 103 can be operated.

[0058] Examples of motions

[0059] Next, an example of the operation of the motion data generation system according to the first embodiment will be described with reference to Figure 6 An example of the operation of the motion data generation system according to the first embodiment will be described. Figure 6 is a diagram showing an example of a method for generating a waist posture in an example of the operation of the motion data generation system according to the first embodiment. In the Figure 6 The flowchart described above, each of the three lanes is assigned to the upper body, the waist, and the lower body.

[0060] The motion capture device 1 captures motion data of the upper body and the waist of the subject H1 (step ST11). The captured motion data includes upper body motion data and captured waist motion data.

[0061] Subsequently, the scaling unit 21 scales the motion data captured in step ST11 to the size of the body of the robot 100 (step ST12).

[0062] In parallel with steps ST11 and ST12, the manually generated motion data acquisition unit 23 acquires lower body motion data and manually generated waist motion data (step ST13).

[0063] Subsequently, the waist motion data generation unit 26 generates waist motion data according to the leg state of the robot 100 (step ST14). Specifically, the leg state determination unit 25 determines the leg state of the robot 100. The waist motion data generation unit 26 generates waist motion data based on the determination result.

[0064] Subsequently, the motion data of the robot 100 is generated (step ST2). Specifically, the motion data of the robot 100 is generated by combining the upper body motion data captured in step ST11, the lower body motion data captured in step ST13, and the waist motion data captured in step ST14.

[0065] In addition, the robot motion control unit 24 can perform stabilization processing on the front-back and left-right direction components of the posture Prt of the robot 100 at the waist 102 according to the target ZMP trajectory to correct the motion data (step ST3). In this case, similar to step ST12, the corrected motion data of the robot 100 is generated by combining the upper body motion data, the lower body motion data, and the waist motion data corrected in step ST3 described above (step ST4).

[0066] Therefore, the motion data of the robot 100 can be generated. The robot motion control unit 24 operates the upper body 101, the waist 102, and the lower body 103 based on the motion data or the corrected motion data.

[0067] The present invention is not limited to the above embodiments and can be appropriately modified without departing from the gist. In addition, the present invention can be implemented by appropriately combining the above embodiments and their examples. The present invention can also be implemented, for example, by causing the CPU 2a to execute for Figure 5 and Figure 6It is implemented by the processed computer program shown. The above program uses various types of non-transitory computer-readable media for storage and can be provided to a computer (including the computer of the information notification device). Non-transitory computer-readable media include various types of tangible storage media. Examples of non-transitory computer-readable media include magnetic recording media (such as floppy disks, magnetic tapes, hard disk drives), magneto-optical recording media (such as magneto-optical discs). In addition, this example includes compact disc-read only memory (CD-ROM), recordable compact disc (CD-R), and rewritable compact disc (CD-R / W). In addition, this example includes semiconductor memories (such as mask read only memory (ROM), programmable read only memory (PROM), erasable programmable read only memory (EPROM), flash ROM, random access memory (RAM)). The program can be provided to the computer through various types of transitory computer-readable media. Examples of transitory computer-readable media include electrical signals, optical signals, and electromagnetic waves. The transitory computer-readable media can provide the program to the computer via a wired communication path such as wires or optical fibers or a wireless communication path.

[0068] In addition, in the above various embodiments, as described in the processing procedure in the motion data generation system 10, the present disclosure can also be in the form of a control method for the motion data generation system 10. In addition. It can be said that the above program is a program for causing the motion data generation system 10 to execute this control method.

Claims

1. A motion data generation system for a robot, the robot including an upper body, a waist, and a lower body, the motion data generation system comprising: A subject motion data acquisition unit that acquires upper body motion data captured from the motion of the upper body of a subject and waist motion data captured from the motion of the waist of the subject; A manually generated motion data acquisition unit that acquires lower body motion data and manually generated waist motion data, where the lower body motion data and the manually generated waist motion data are generated by manual input of a user; and A robot motion control unit that includes a leg state determination unit and a waist motion data generation unit, where: The leg state determination unit determines whether at least one leg of the robot is in a swinging state or whether both legs of the robot are in a standing state; When the leg state determination unit determines that at least one leg of the robot is in the swinging state, the waist motion data generation unit generates the manually generated waist motion data as the robot waist motion data; When the leg state determination unit determines that both legs of the robot are in the standing state, the waist motion data generation unit generates the captured waist motion data as the robot waist motion data; and The robot motion control unit controls the upper body, the waist, and the lower body of the robot based on the upper body motion data, the robot waist motion data, and the lower body motion data, respectively; Wherein, during a transition period from a standing phase in which both legs of the robot are in the standing state to a swinging phase in which at least one leg of the robot is in the swinging state, the waist motion data generation unit generates the robot waist motion data by synthesizing the manually generated waist motion data and the captured waist motion data.

2. The motion data generation system for a robot according to claim 1, wherein, From the start to the end of the transition period, in the robot waist motion data generated from the synthesized data, the reference ratio referring to the manually generated waist motion data gradually increases with respect to the captured waist motion data.

3. A motion data generation method for a robot, which is executed in the motion data generation system of the robot, the robot including an upper body, a waist, and a lower body, the motion data generation method comprising: Steps of acquiring upper body motion data captured from the motion of the upper body of a subject and waist motion data captured from the motion of the waist of the subject; Steps of acquiring lower body motion data and manually generated waist motion data, where the lower body motion data and the manually generated waist motion data are generated by manual input of a user; Steps of determining whether at least one leg of the robot is in a swinging state or whether both legs of the robot are in a standing state; Steps of, when it is determined that at least one leg of the robot is in the swinging state, generating the manually generated waist motion data as the robot waist motion data, and when it is determined that both legs of the robot are in the standing state, generating the captured waist motion data as the robot waist motion data; Steps of controlling the upper body, the waist, and the lower body of the robot based on the upper body motion data, the robot waist motion data, and the lower body motion data, respectively; and During a transition period from a standing phase in which the two legs of the robot are in the standing state to a swinging phase in which at least one leg of the robot is in the swinging state, a step of generating the robot waist motion data by synthesizing the manually generated waist motion data and the captured waist motion data.

4. A storage medium storing a motion data generation program for a robot, the motion data generation program being executed by a computer that operates as an arithmetic device in the motion data generation system of the robot, the robot including an upper body, a waist, and a lower body, the motion data generation program causing the computer to execute: Steps of acquiring upper body motion data captured from the motion of the upper body of the subject and waist motion data captured from the motion of the waist of the subject; Steps of acquiring lower body motion data and manually generated waist motion data, the lower body motion data and the manually generated waist motion data being generated by manual input of a user; Steps of determining whether at least one leg of the robot is in a swinging state or whether both legs of the robot are in a standing state; Steps of generating the manually generated waist motion data as the robot waist motion data when it is determined that at least one leg of the robot is in the swinging state, and generating the captured waist motion data as the robot waist motion data when it is determined that both legs of the robot are in the standing state; Steps of controlling the upper body, the waist, and the lower body of the robot based on the upper body motion data, the robot waist motion data, and the lower body motion data respectively; and Steps of generating the robot waist motion data by synthesizing the manually generated waist motion data and the captured waist motion data during a transition period for transitioning from a standing phase where the two legs of the robot are in the standing state to a swinging phase where at least one leg of the robot is in the swinging state.

Images